longer food chains and crowded niche space: effects of

Longer food chains and crowded niche space:effects of multiple invaders on desert stream foodweb structureTimothy E. Walsworth1,*, Phaedra Budy1,2, Gary P. Thiede11Department of Watershed Sciences, Utah State University, Logan, UT 84322, US2U.S. Geological Survey Utah Cooperative Fish and Wildlife Research Unit, Utah State University, Logan, UT 84322, US

Accepted for publication January 19, 2013

Abstract – Tributaries of the Colorado River Basin, historically home to a complex of endemic omnivorescollectively referred to as the ‘three species’; flannelmouth sucker (Catostomus latipinnis), bluehead sucker(C. discobolus) and roundtail chub (Gila robusta), have experienced the establishment of numerous non-native fishspecies. In this study, we examine the impacts of the trophic ecology of non-native fishes on the ‘three species’ inthe San Rafael River, Utah, USA. We employ a suite of abundance comparisons, stable isotope techniques andsize-at-age back-calculation analyses to compare food web structure and growth rates of the ‘three species’ in studyareas with and without established populations of non-native species. We found that the ‘three species’ are moreabundant in areas with few non-native fishes present, regardless of habitat complexity. Stable isotope analysesindicate non-native fishes lengthen the food chain by 0.5 trophic positions. Further, the trophic niche spaces of thenative fishes shift and are narrower in the presence of non-native fishes, as several non-native species’ trophic nichespaces overlap almost entirely with each of the ‘three species’ (bluehead sucker and flannelmouth sucker 100%,roundtail chub 98.5%) indicating strong potential for competition. However, the ‘three species’ demonstrated noevidence of reduced growth in the presence of these non-native fishes. Collectively, these results suggest that whilenon-native fishes alter the food web structure presenting novel sources of predation and competition, mechanismsother than competition are controlling the size-structure of ‘three species’ populations in the San Rafael River.

Key words: non-native species; competition; novel predators; growth; stable isotope analysis

Introduction

The impact of human development increasinglythreatens the native biodiversity of freshwater ecosys-tems (Ricciardi & Rasmussen 1999; Dudgeon et al.2006). Anthropogenic alterations to rivers result inphysical (e.g., flow alteration, sedimentation, chan-nelisation), chemical (e.g., pollution, nutrient loading)and biological (e.g., invasive species, disease) degra-dation of natural habitats. These alterations canfavour novel life-history strategies (Olden et al.2006), decrease resource availability (Tyus & Karp1990; Brouder 2001), alter the flow of energythrough ecosystems (Sousa et al. 2008) and increase

predation and competition pressure on native species(Tyus & Saunders 2000). As such, anthropogenicalterations can result in new evolutionary andcontemporary pressures on native biota that evolvedunder a specific suite of environmental conditions,often dramatically restructuring communities.Biotic communities are structured not only by the

environmental tolerances of the species present, butalso by interactions among the species of the commu-nity. Interspecific competition for resources can limitgrowth and survival of inferior competitors (e.g.,Schoener 1983) and thus limit the distribution andabundance of species. However, negative population-level effects resulting from competition have proven

Correspondence: Timothy E. Walsworth, Department of Watershed Sciences, Utah State University, Logan, UT 84322. E-mail: [email protected]*Current address: School of Aquatic and Fishery Sciences, University of Washington,Box 355020, Seattle WA 98195

doi: 10.1111/eff.12038 1

Ecology of Freshwater Fish 2013 Published 2013. This article is a U.S.Government work and is in the public domain in the USA.

ECOLOGY OFFRESHWATER FISH

difficult to detect in many, though not all, studies(reviewed in Connell 1983). In addition, predatorspecies can have a particularly strong role in shapingcommunity structure through both direct (e.g., Kitc-hell et al. 1997) and indirect pathways (e.g., Werneret al. 1983; Romare & Hansson 2003). Alterations totrophic structure, such as the establishment of non-native species, can have significant impacts on com-munity structure (e.g., Reissig et al. 2006). However,the impact of invasive predators and competitors isnot consistent for all recipient systems and is influ-enced by ecosystem complexity (Case 1990; Carey &Wahl 2010, 2011) and evolutionary history (Kitchellet al. 1997; Shea & Chesson 2002).Rivers and streams of the Colorado River Basin,

home to a historically depauperate and highly ende-mic fish fauna, have been particularly affected byanthropogenic disturbances (Minckley & Deacon1968), due in part to the region’s arid climate andrapidly expanding human population. While degrada-tion of the physical environment is often argued tobe the primary cause of population and range reduc-tions of native fishes (Ross 1986), interactions withnon-native fishes may be equally or even more influ-ential than habitat loss in the extinction of species(Tyus & Saunders 2000; Woodford 2009a). Invasionby non-native fishes has the potential to alter theenergetic pathways of the historic food web, resultingin significant and negative ecosystem impacts, suchas modified habitat coupling, nutrient cycling ratesand ecosystem resilience (Eby et al. 2006; Brittonet al. 2010; Pilger et al. 2010). Non-native speciesoften maintain a competitive advantage over nativefishes (Shea & Chesson 2002; Cox & Lima 2006)and can have particularly strong impacts on simplefood webs (Carey & Wahl 2010), such as those ofthe historical Colorado River Basin (Tyus & Nikirk1990; Tyus & Saunders 2000; Unmack & Fagan2004).In addition, the interaction between a degraded

physical habitat and non-native species presence canresult in complex synergistic effects on native speciesand their habitat (Brook et al. 2008). Physical habitatdegradation, for example, can aid in the establish-ment and dispersal of invasive species (Marvier et al.2004). While change in flow regimes may decouplethe life histories of native fishes from their currentenvironmental template, many invasive fishes arepre-adapted to the new conditions (Olden et al. 2006;Johnson et al. 2008). In addition , the impact of non-native fishes has been suggested to be strongest attimes of low flow (Pilger et al. 2010), a nearlyperpetual state in degraded desert rivers. This combi-nation of physical degradation and biotic invasions inthe Colorado River Basin has contributed to the extir-pation and the federal listing of many endemic fishes

under the Endangered Species Act (Minckley &Deacon 1968; USDI, Fish & Wildlife Service 1994)and the protection of many others under conservationagreements between states of the Colorado RiverBasin (UDWR, Utah Division of Wildlife Resources2006).The bluehead sucker (Catostomus discobolus),

flannelmouth sucker (C. latipinnis) and roundtailchub (Gila robusta) represent a highly imperilledecological complex of fishes, hereafter, collectivelythe ‘three species’, native to the upper ColoradoRiver Basin, U.S.A. These endemic fishes are listedas state species of concern and subject to a range-wide conservation agreement aimed at ensuring theirpersistence throughout their range (UDWR 2006).Many of the non-native fishes present in the basinhave the potential to negatively impact populations ofthese native fishes through competition or predation(e.g., Bestgen & Propst 1989; Tyus & Saunders2000; Johnson et al. 2008). In fact, predation onearly life stages of native fish by non-native fish hasbeen implicated in total recruitment failure for nativefishes in the Colorado River Basin and other systems(Meffe 1985; Woodford 2009b).Populations of imperilled species may be able to

persist in an area if recruitment cannot balance withmortality, provided there is a source habitat capableof supplying colonisers to the sink habitat (Pulliam1988). Previous study has suggested that populationsof the ‘three species’ in the downstream, degradedportions of the San Rafael River, a tributary of theGreen River, occupy a sink habitat, and are main-tained through colonisation from sources in both theupper portions of the San Rafael River and themain-stem Green River (e.g., Bottcher 2009). Identi-fication of the mechanisms (i.e., competition and/orpredation) by which non-native fishes interact withthe ‘three species’ may illuminate the causes ofsource-sink structure. In this article, we address thefollowing questions: (i) do non-native fishes occupypiscivorous trophic positions and/or change the tro-phic position occupied by native fishes and (ii) doesthe presence of non-native fishes alter the growth ofnative fishes? In addressing these questions, weexplore a combination of food web and growth anal-yses among patches of a river with alternate inva-sion histories to advance our understanding of howtrophic interactions impact community compositionand population structure.

Study area

The San Rafael River (hereafter ‘SRR’), Utah,U.S.A., is an ideal study stream as it has experienceddegradation representative of many tributaries of theColorado River Basin (i.e., altered hydrograph, water

2

Walsworth et al.

loss, channelisation, homogenisation of habitat, inva-sive species establishment; Walker & Hudson 2004;Bottcher 2009). The SRR drains 4500 km2 of south-eastern Utah and is formed by the confluence ofFerron, Cottonwood, and Huntington creeks (Fig. 1).The SRR is a spring snowmelt and autumn monsoondriven system, and flows approximately 175 km fromits headwaters in the Manti–La Sal National Forest toits confluence with the Green River near the town ofGreen River, Utah. The SRR is one of the most over-allocated rivers in Utah (Walker & Hudson 2004)and the lower 64 km are frequently dewatered duringthe summer irrigation season (Bottcher 2009).Native fish species currently occupying the SRR

for at least a portion of their life history include theflannelmouth sucker, bluehead sucker, roundtailchub, speckled dace (Rhinichthys osculus), as well asoccasional, transient bonytail (Gila elegans), razor-back sucker (Xyrauchen texanus) and Coloradopikeminnow (Ptychocheilus lucius). Non-native spe-cies present include red shiner (Cyprinella lutrensis),sand shiner (Notropis stramineus), fathead minnow(Pimephales promelas), channel catfish (Ictaluruspunctatus), black bullhead (Ameiurus melas), com-mon carp (Cyprinus carpio), white sucker (Catosto-mus commersoni), green sunfish (Lepomis cyanellus)and virile crayfish (Orconectes virilis).

In this study, we delineate the upper and lower SRRby the San Rafael Reef, dividing the river into the sec-tion flowing through the San Rafael Swell, character-ised by deep canyons (upper SRR), and the sectionflowing through the San Rafael Desert, characterisedby primarily open desert terrain (lower SRR). Naturaland anthropogenic barriers in the San Rafael Reef andat the Hatt Ranch diversion dam limit upstream move-ment of fishes. Twelve 300 m long sampling siteswere established in total (Fig. 1). The sites in theupper SRR (N = 4) were chosen opportunistically dueto extremely rugged terrain, while sites in the lowerSRR (N = 8) were chosen by a systematic sampledesign with random seed start (Bottcher 2009).

Methods

Collection of biotic community data

Sampling events occurred during the spring (i.e., priorto spring snowmelt run-off; April and May), summer(i.e., immediately after recession of spring runoff;June and July) and autumn (i.e., October) of 2010.Fish from all habitat types were sampled via canoeelectrofishing, and all native and non-native fisheswere identified to species, anesthetised with tricainemethanosulphate (MS-222), weighed, measured fortotal length and clipped (lower caudal fin) for stableisotope analysis. Tissue samples were preserved andstored in 95% ethanol until processing later in the lab,and native fishes were fin-clipped for age determina-tion (see below). Storing tissue samples in ethanolhas been shown to enrich d13C signatures in somestudies (Kaehler & Pakhomov 2001; Kelly et al.2006), to have no significant effect in other studies(Sarakinos et al. 2002; Serrano et al. 2008) and tocause small shifts relative to ecological variation(Kelly et al. 2006). We were confident using thesetissues given that all indices were relative and as longas carbon signatures were interpreted with caution.Following sample collection, native fishes were heldin a recovery tank until they showed no effects of theanaesthetic before being returned to the river.Benthic invertebrates were sampled at each food

web site with a Surber sampler. For each samplingevent, eight 0.09 m2 samples were taken from rifflesin each site. Samples within a site and date werecombined in 95% ethanol, returned to the lab wherethey were sorted, identified to family (or lower) andfunctional feeding group, and counted.

Relative abundance of native fishes in presence ofnon-native species

The sampling sites were placed into three categoriesof degradation (Fig. 1). The high quality category

Fig. 1. Map of the San Rafael River watershed. Inset showslocation of watershed (cross-hatched) in Utah, USA.

3

Effects of multiple invaders on imperiled desert fishes

(all located in the upper SRR) had widely availablecomplex habitat (i.e., riffles, pools and backwaters)and very low to zero densities of non-native fishespresent. The medium quality category (all located inthe lower SRR, with one site upstream of the HattRanch diversion dam, yet downstream of the SanRafael Reef) had widely available complex habitat,but high densities of non-native fishes present. Thepoor quality category (all in the lower SRR) all hadvery limited complex habitat available and high den-sities of non-native fishes. Habitat complexity wascalculated as the per cent of the reach consisting ofriffle, pool or backwater habitat as determined byin-stream habitat surveys (Walsworth 2011). Whilethe surrounding terrain is somewhat different betweenthe upper and lower SRR (Fortney et al. 2011), thein-stream habitat in high and medium quality reacheswas similar. A repeated measures analysis of vari-ance, repeated on sampling site, was run to comparethe catch-per-unit-effort (CPUE) of native fishes andof non-native fishes between sites in the differentquality categories.

Stable isotope analysis of food web structure

Once returned to the lab, caudal fin clips and benthicinvertebrates of the collector-filterer functional feed-ing groups (most abundant) were prepared for stableisotope analysis. Tissue samples were dried for 48 hat 70 °C, crushed into a homogeneous powder withmortar and pestle and placed in tin capsules for ship-ment. To examine the effect of non-native species onthe trophic structure of the SRR, samples from theupper SRR (high quality sites) and from the lowerSRR (medium and poor quality sites) were analysedfor stable isotopic signatures separately. We chosethis division due to the results of our repeated mea-sures ANOVA, as it separated areas of the river with(lower SRR) and without (upper SRR) abundantnon-native fishes. When possible, we collected aminimum of 10 samples from each species in boththe upper and lower SRR (Vinson & Budy 2011).Capsules were sent to the Washington State Univer-sity Stable Isotope Core Lab (Pullman, WA) fornatural abundance analyses of 13C and 15N. Isotopicsignatures are reported in d-notation:

d13C or d15N ¼ Rsample

Rstandard

� �� 1

� ��1000;

where R is the ratio of 13C/12C or 15N/14N. The stan-dard for d13C is PeeDee belemnite and for d15N isatmospheric nitrogen. The stable isotopic signaturesof the collector-filterer invertebrate group were usedas a baseline to allow comparison between different

sites along the longitudinal gradient of the SRR, asthis was the only functional group of invertebratesconsistently collected in Surber samples. Collector-filterer insects accounted for a median of 88.6%(mean = 76.4%; SE = 3.7%; Skewness = �1.47) ofSurber sample composition. The primary collector-fil-terer families collected and analysed in this studywere Hydropsychidae (Order: Trichoptera) and Simu-liidae (Order: Diptera). Families were analysed forstable isotope signature separately, and the resultswere combined to measure the mean collector-filterersignature. Each of these families was found in boththe upper and lower SRR.Carbon and nitrogen isotope signatures were cor-

rected for basal resource variation before further anal-ysis. Carbon signatures were corrected using thefollowing equation:

d13Ccor ¼ d13Ci � d13Cmeancf

CRcf

where d13Ci is the carbon isotope signature of organ-ism i, d13Cmeancf is the mean collector filterer carbonisotopic signature from the site where organism i wascollected and CRcf is the carbon range of the samecollector filterer invertebrates as above (Olsson et al.2009). Nitrogen values were corrected (d15Ncor) forbaseline variation by:

d15Ncorj;i ¼ d15Nj;i � d15Ncf ;i � d15Ncf ;min� �

where d15Nj,i is the nitrogen signature of sample j at

site i, d15Ncf ;i is the mean d15N of collector-filterers

at site i, and d15Ncf ;min is the minimum mean collec-tor-filterer d15N of all sites (adapted from Cabana &Rasmussen 1996 to display stable isotope signaturesin d-space). Although baseline signatures differedbetween individual sampling sites (d13C SD = 1.05,d15N SD = 1.1), we did not detect any longitudinalpatterns in carbon or nitrogen baselines. Due to heter-ogeneous catch rates associated with their low abun-dance and imperilled status, a subsample ofindividuals captured at sites with high catch rateswere analysed for stable isotope signatures to cha-racterise the food web structure throughout the upperand lower SRR. Due to limited sample sizes for the‘three species’, especially in the lower SRR, stableisotope data were pooled across seasons (spring, sum-mer, and autumn).Species’ trophic niche space was analysed using

methods presented in Layman et al. (2007). For eachspecies, the following measures were calculated; tro-phic position (TP), nitrogen range (NR), carbon range

4

Walsworth et al.

(CR), trophic niche width, trophic niche overlap andfood chain length. Trophic position was calculatedusing the following equation:

TPi ¼ d15Ncori � d15Ncorcf3:4

þ 2;

where TPi is the trophic position of species i, Ncori isthe corrected nitrogen signature of species i, andNcorcf is the corrected nitrogen signature of collectorfilterer invertebrates (Vander Zanden & Rasmussen1999). Collector-filterer invertebrates were assumed tohave a trophic position of 2 and we assumed trophicfractionation of d15N to be 3.4& as trophic levelincreases (Minagawa & Wada 1984). Trophic positionand d13Ccor signatures for each of the ‘three species’were compared between the upper and lower SRRwith Student’s t-tests. Nitrogen range and carbonrange were calculated from the following equations:

NR ¼ d15Ncormax � d15Ncormin;

CR ¼ d13Ccormax � d13Ccormin

for each species (Layman et al. 2007). Niche widthwas calculated as the convex hull area of isotopic sig-natures of each species plotted in C-N bivariate space(Layman et al. 2007). Niche overlap was calculatedas the per cent of a species’ niche width area that isoverlapped by the niche width area of another speciesin the food web. Food chain length was calculated asthe highest TP detected in the upper or lower SRR.

Calculation of native fish growth rates

The basal portion of the second dorsal fin ray wascollected from each individual of the ‘three species’captured during sampling, and stored in vials untilreturned to the lab. The fin rays were placed in moldsand heated at 70 °C for 12 h to allow the mold toharden. The molds were then mounted on metalchucks and cut using a Buehler Isomet low-speedsaw (Buehler Ltd., Lake Bluff, IL, USA). The freshedge was sanded, polished and a second cut wasmade to produce a thin (~1.5 mm) section. The sec-tion was placed on a microscope slide and polished ifnecessary. Fin ray sections were placed under a cam-era-mounted microscope and back-lit to reveal annuli.Digital images of each section were captured andanalysed for length-at-age using the Frasier-Leeback-calculation:

Lt ¼ cþ LT � cð Þ StST

� �

where Lt is the length of the fish at time t, LT is thetotal length of the fish, c is the length of the fish at

age 1, St is the length of the fin ray at time t and STis the total length of the fin ray (Francis 1990). Thelength of each species at age 1 (c) was estimatedfrom length-frequency plots. Length-at-age data foreach species were then fit to a repeated measuresnon-linear Von Bertalanffy growth curve (VBGC;Jones 2000):

Lt ¼ L1 1� e�k t � t0ð Þ�

where L∞ represents the maximum possible lengthachieved by the fish, Lt is the fish’s length at time t,t0 is the theoretical age of the fish when its lengthequals zero and k is the Brody growth rate coefficient(Ricker 1975), a measure of the rate of approach toL∞. Growth rates of fish in the upper and lower SRRwere examined by comparing the standardised residu-als from the VBGC of individual fish captured ineach section. A Student’s t-test was used to comparethe residuals between groups. All statistical analyseswere run in the R Statistical Environment (R Devel-opment Core Team 2011) with an a priori a of 0.05.

Results

Relative abundance

The lower SRR fish community species compositionwas dominated by non-native fishes, while the fishcollections in the upper SRR were almost entirelycomposed of native fishes (Table 1). In addition,juvenile individuals of the ‘three species’ wereuncommon in the electrofishing samples (Table 1).The ‘three species’ were captured in significantlygreater numbers in the high-quality sites (CPUE= 26.22 � 3.69 fish per hour [mean � 1.96∙SE])than in either the medium (CPUE = 3.93 � 1.14 fishper hour) or low-quality sites (CPUE = 4.30 � 1.04fish per hour; repeated measures ANOVA, P = 0.001;Fig. 2a), while non-native fishes were captured insignificantly greater numbers in both the medium(CPUE = 61.01 � 11.60 fish per hour) and low-(CPUE = 53.09 � 6.13 fish per hour) quality sitesthan in the high-quality sites (CPUE = 0.79 � 0.46fish per hour; repeated measures ANOVA P < 0.001;Fig. 2b). No significant seasonal or interaction effectswere detected for CPUE of either native or non-native fishes (all repeated measures ANOVA P > 0.05).The only non-native fishes captured upstream of theSan Rafael Reef were 2 green sunfish and 1 Utahchub (Gila atraria). Non-native white sucker werecaptured in the lower SRR for the first time in 2010.The relative abundances of benthic invertebrates

were not significantly different between the upperand lower SRR (ANOVA, P = 0.432). When compar-

5


ing only densities of collector filterer insects only,there was still no significant difference between theupper and lower SRR (ANOVA, P = 0.319).

Stable isotope analysis

The upper SRR can be broadly characterised by ashort food chain (food chain length = 3.51; Table 2,Fig. 3a) and food web populated by few species. Thebluehead sucker (‘BHS’, TP = 2.75 � 0.18[mean � 1.96∙SE]), flannelmouth sucker (‘FMS’,TP = 2.95 � 0.22) and roundtail chub (‘RTC’,TP = 2.80 � 0.24) occupied similar trophic positions(ANOVA P = 0.38) that signify the assimilation of ben-thic invertebrates. The ‘three species’ displayed sig-nificantly different d13Ccor signatures (BHS d13Ccor

= �0.65 � 0.18; FMS d13Ccor = 0.77 � 0.24; RTCd13Ccor = 1.91 � 0.40; ANOVA P < 0.001). The blue-head sucker was more depleted in carbon than eitherthe flannelmouth sucker (t-test P < 0.001) or theroundtail chub (t-test P < 0.001), and the flannel-mouth sucker was more depleted in carbon signaturethan the roundtail chub (t-test P < 0.001). Speckleddace occupied the highest mean trophic position inthe upper SRR (TP = 3.18 � 0.08) and demonstratedan intermediate d13C signature among the upper SRRfish community (d13Ccor = 0.64 � 0.59).While the food web of the upper SRR was struc-

turally simple, the food web of the lower SRR waspopulated by more species distributed across abroader range of trophic positions (food chainlength = 4.01; Table 2, Fig. 3b). As in the upperSRR, the bluehead sucker (TP = 2.82 � 0.06) andflannelmouth sucker (TP = 2.84 � 0.06) occupied

Table 1. Number and mean, minimum (‘Min’) and maximum (‘Max’) total lengths of individuals sampled in the San Rafael River by species and location. Cellswith no data available marked with ‘–’.

Species

Upper San Rafael River Lower San Rafael River

NMeanLength

MinLength

MaxLength N

MeanLength

MinLength

MaxLength

Native SpeciesAge-0 sucker 18 26.7 21 38 37 25.9 18 32Bluehead sucker (BHS) 155 178.9 75 290 16 214.0 176 260Flannelmouth sucker (FMS) 110 284.7 106 426 25 191.8 117 325Roundtail chub (RTC) 40 145.2 55 261 21 180.2 74 290Speckled dace (SDD) 321 63.9 29 98 15 70.9 60 97

Non-native SpeciesBlack bullhead (BBH) 0 – – – 7 188.6 140 235Channel catfish (CLC) 0 – – – 16 248.3 60 384Common carp (CNC) 0 – – – 6 451.0 407 511Fathead minnow (FHM) 0 – – – 28 66.0 54 74Green sunfish (GNS) 2 95.5 51 140 15 118.9 69 165Red shiner (RDS) 0 – – – 453 55.8 20 84Sand shiner (SDS) 0 – – – 711 52.6 15 94Utah chub 1 135.0 135 135 0 – – –White sucker (WES) 0 – – – 4 166.8 120 220

High Medium Poor0

10

20

30

40

50

60

Nat

ive

Spec

ies

CPU

E (

fish

hr–

1 )

Reach quality

High Medium Poor0

25

50

75

100

125

150

Non

-nat

ive

Spec

ies

CPU

E (

fish

hr–

1 )

Reach quality

(a)

(b)

Fig. 2. Catch-per-unit-effort (fish/hr) of native (a) and non-native(b) fish in the San Rafael River by site quality. The high-qualitysites have significantly greater CPUE of native fishes (ANOVAP = 0.001) and significantly lower CPUE of non-native fishes(ANOVA P < 0.001) than either the medium or poor quality sites.

6

Walsworth et al.

similar trophic positions (t-test P = 0.70), and thetwo species demonstrated significantly different car-bon signatures (BHS d13Ccor = 1.26 � 0.18; FMSd13Ccor = 2.02 � 0.37; t-test P = 0.002). The round-tail chub had a significantly elevated trophic position(TP = 3.34 � 0.27) in the lower SRR compared to

the two native suckers (ANOVA P < 0.001), and wassignificantly elevated from its trophic position in theupper SRR (t-test P = 0.009). The bluehead suckerand flannelmouth sucker carbon signatures were sig-nificantly enriched in the lower SRR compared to theupper SRR (both t-tests P < 0.001). All remainingstable isotope measures for the ‘three species’ weresimilar between the upper and lower SRR (t-tests, allP > 0.05). Green sunfish held the highest trophicposition in the lower SRR (TP = 3.54 � 0.24), andfive other non-native species occupied higher trophicpositions than at least two of the ‘three species’.The stable isotope signatures of fishes in the upper

SRR spanned a broad array of carbon sources andtrophic positions (CR = 3.71, NR = 4.08, nichewidth = 12.79). Bluehead sucker occupied a narrowtrophic niche (CR = 0.53, NR = 3.37, nichewidth = 6.96; Fig. 4a), aligning closely with collec-tor-filterer insects with respect to carbon signature.The bluehead sucker niche space did not overlap withother native species trophic niches.. Flannelmouthsucker had a slightly wider, less distinct, trophicniche space than the bluehead sucker (FMSCR = 1.27, NR = 3.24, niche width = 7.14, over-lap = 22.4%). The roundtail chub occupied the wid-est trophic niche space out of the ‘three species’ inthe upper SRR, and had less niche overlap than theflannelmouth sucker (CR = 1.89, NR = 3.32, nichewidth = 8.41, overlap = 6.9%).The stable isotope signatures of fishes in the lower

SRR was similar to that used in the upper SRR,although a slightly wider range of trophic positionswere inhabited and there was greater variation inresource use at all given trophic positions(CR = 5.15, NR = 4.96, niche width = 18.12;Fig. 4b). Bluehead sucker trophic niche space in the

Table 2. Stable isotope signatures by species and location in the San Rafael River. Sample size (N), mean corrected stable isotope signature values andstandard errors are reported. Cells with no data available marked with ‘–’.

Species

Upper San Rafael River Lower San Rafael River

N d13Ccor (SE) d15Ncor (SE) N d13Ccor (SE) d15Ncor (SE)

Native SpeciesCollector-Filterer Insects(ColFil) 18 0.00 (0.09) 3.53 (0.12) 81 0.00 (0.04) 3.53 (0.06)Bluehead sucker (BHS) 10 �0.53 (0.06) 6.14 (0.36) 10 1.26 (0.09) 6.33 (0.12)Flannelmouth sucker (FMS) 10 0.77 (0.12) 6.74 (0.37) 11 2.02 (0.19) 6.39 (0.12)Roundtail chub (RTC) 9 1.91 (0.20) 6.26 (0.39) 10 2.38 (0.31) 7.78 (0.41)Speckled dace (SDD) 5 0.64 (0.30) 7.54 (0.14) 0 – –

Non-native SpeciesBlack bullhead (BBH) 0 – – 4 1.84 (0.47) 8.45 (0.33)Channel catfish (CLC) 0 – – 10 2.68 (0.35) 8.32 (0.23)Common carp (CNC) 0 – – 5 1.47 (0.39) 5.84 (0.68)Fathead minnow (FHM) 0 – – 7 0.90 (0.07) 7.53 (0.43)Green sunfish (GNS) 0 – – 7 1.81 (0.25) 8.79 (0.42)Red shiner (RDS) 0 – – 37 1.04 (0.15) 6.89 (0.10)Sand shiner (SDS) 0 – – 28 1.82 (0.23) 7.37 (0.13)White sucker (WES) 0 – – 4 2.01 (0.67) 6.27 (0.38)

●

−2 −1 0 1 2 3 4

−2 −1 0 1 2 3 4

4

6

8

10

δ15N

cor

δ13Ccor

ColFil

●SDD

●FMS●

BHS ●RTC

●4

6

8

10

δ15N

cor

δ13Ccor

ColFil

●FMS

●BHS

●RTC

●CLC●

GNS●

BKB

●WES

●RDS ●SDS

●CNC

●FHM

(a)

(b)

Fig. 3. Stable isotope signatures (mean values � 1.96 SE) andfood web structure of the upper (a) and lower (b) San RafaelRiver. The ‘three species’ icons are enlarged. Mean values � 1.96SE shown. See Table 2 for abbreviations.

7


lower SRR was much narrower and less distinct thanin the upper SRR (CR = 0.77, NR = 1.21, nichewidth = 3.25, overlap = 100%). Flannelmouth suckertrophic niche space was also considerably more nar-row and less distinct than in the upper SRR(CR = 1.89, NR = 1.22, niche width = 5.03, over-lap = 100%). Roundtail chub trophic niche spacewas slightly wider in the lower SRR (CR = 2.95,NR = 3.69, niche width = 9.89), but was overlappedwith other species’ trophic niche spaces to a muchgreater degree (overlap = 98.5%) than in the upperSRR. Red shiners, sand shiners, channel catfish, com-mon carp and white suckers each occupied a trophicniche that overlapped substantially with those of the‘three species’.

Growth

The oldest bluehead suckers (age = 6 year) were cap-tured in the upper SRR, while those captured in thelower SRR had a greater mean age than thosecaptured in the upper SRR (lower SRR mean

age = 4.27 year, upper SRR mean age = 4.00 year).Flannelmouth sucker and roundtail chub attained thesame maximum age measured in the SRR (maximumage = 8 year). Flannelmouth sucker in the upperSRR exhibited an older maximum age (8 year) mea-sured, as well as an older mean age (5.07 year) thanthose aged in the lower SRR (maximum = 7 year;mean = 4.39 year). The lower SRR exhibited boththe oldest individual (8 + year) and a greater meanroundtail chub age (4.59 year) than the upper SRR(4.20 year). More individuals were analysed for ageand growth than were analysed for stable isotope sig-nature (Upper SRR BHS N = 43, FMS N = 31, RTCN = 10; Lower SRR: BHS N = 15, FMS N = 23,RTC N = 17).Flannelmouth sucker demonstrated the greatest

model-estimated maximum length (L∞ = 481.46 mm;Fig. 5) and roundtail chub the smallest model-esti-mated maximum length (L∞ = 267.00 mm) of the‘three species’. Bluehead sucker exhibited the great-est Brody growth rate coefficient (k = 0.21) and flan-nelmouth sucker the lowest predicted Brody growthrate coefficient (k = 0.19). No differences in length-at-age were detected between fish captured in theupper and lower SRR for either flannelmouth sucker(t-test P = 0.72) or roundtail chub (t-test P = 0.48).However, bluehead sucker captured in the lower SRRwere significantly larger at age than those captured inthe upper SRR (t-test P = 0.012).

−2 0 2 4

2

4

6

8

10

12

δ15N

cor

δ13Ccor

−2 0 2 4δ13Ccor

2

4

6

8

10

12

δ15N

cor

(a)

(b)

Fig. 4. Trophic niche spaces occupied by fish species in theupper (a) and lower (b) San Rafael River. The ‘three species’ areshown with solid lines. Solid lines: blue = bluehead sucker,green = flannelmouth sucker and yellow = roundtail chub;Dashed lines: blue = sand shiner, red = red shiner, grey = chan-nel catfish, yellow = green sunfish, light blue = fathead minnow,black = collector filterer insects; Dotted lines: yellow = commoncarp, red = white sucker, black = black bullhead; Dash-dottedlines; red = speckled dace.

●●●

●● ● ● ●

●●

●● ●

●

●

●●

●

●●

● ●

●●

●

● ●

●●

● ●

●

●●

●● ●

●

●

●●

● ● ●

●

●●

●●

●● ● ●

●

●●

●●

●

●

●●

● ●

●●

●●

●● ●

●●

●

●●

●

●●

●●

● ● ● ● ● ●●

●●

●

●●

●

●

●●

●

●●

●●

●

●●

●

●●

●

●●

●

●●

● ●●

● ●

0100200300400500600

●●

●●

●●

● ●●●●●

●

●

●●

●●

●●

●

●●

●● ●

●● ●

●●

● ●

●●

●

●●● ●

●● ●

●

●

●●

●●

●

●●

● ●● ●●

●● ●

●●

●

●●

●●

●

●● ●

●●

●

●●

●

●●

●

●● ● ●

●● ●

●

●●

●●

●

●●

●

●● ●

●● ●

●● ● ●

●●

●

●●

● ●

●● ● ●

●

● ●●

●●

●●

050

100150200250300350

Bac

kcal

cula

ted

leng

th (m

m)

●

●●

●●

●

●●

●

●●

●

●

●●

●

●●

● ●

●● ●●

●●

●

●●

●●

●

050

100150200250300

0 1 2 3 4 5 6 7 8Age

(a)

(b)

(c)

Fig. 5. Size-at-age with Von Bertalanffy growth function for flan-nelmouth sucker (a), bluehead sucker (b), and roundtail chub (c)in the San Rafael River. Black triangles represent fish captured inthe lower San Rafael River; grey circles represent fishes capturedin the upper San Rafael River. Note different y-axes.

8

Walsworth et al.

Discussion

In this study, we analysed stable isotope signaturesand growth (a fitness-related measure) to examinethe effect of non-native species on trophic status andgrowth of populations of imperilled desert fishes.Our analysis of catch rates between areas with andwithout complex habitat and non-native fishesrevealed that areas of the San Rafael River withoutnon-native fishes held significantly greater abun-dances of the ‘three species’ than those areas withnon-native species present, regardless of habitat qual-ity. Results of the stable isotope analysis revealedthat non-native fish species have significantly alteredfood web structure in the San Rafael River. Afterdetecting these changes to the trophic structure ofthe river, we examined the relative growth and con-dition of the ‘three species’ to determine whether theestablishment of the non-native fish populations hashad an effect on the relative fitness of theseimperilled native fishes. Greater understanding ofthese issues is critical to the effective conservationand management of these endemic species, and hasimplications for the conservation of species threa-tened by habitat loss and non-native species intro-ductions worldwide.Non-native species can alter the flow of nutrients

and energy through an ecosystem through alterationof the number and strength of trophic linkages in afood web (Eby et al. 2006; Sousa et al. 2008; Brittonet al. 2010). In the process, they may also facilitatethe establishment of other non-native species (Sim-berloff & Von Holle 1999). The stable isotope signa-tures of the ‘three species’ detected in this studyindicate trophic levels consistent with the dietsobserved in previous studies (Childs et al. 1998; Bez-zerides & Bestgen 2002; Quist et al. 2006). Theestablishment of non-native fishes has increased thenumber of resident fish species in the food web fromfour in the upper San Rafael River (and assumed pre-invasion lower San Rafael River) to eleven (includingseven non-native) species in the lower San RafaelRiver. Although the food web of the upper SanRafael River consists primarily of a few native, sec-ondary consumers (the ‘three species’ and speckleddace), the food web of the lower San Rafael Riverconsists of a crowded level of secondary consumers,as well as a diverse level of tertiary consumers.Acknowledging that the potential ecological impactof an invading species is context dependent (i.e., onthe abiotic conditions and biotic community of theinvaded ecosystem; Ruesink 2003), the high diversityof invading species in the San Rafael River increasesthe likelihood that at least one of the establishednon-native species will be a high-impact invader(Ricciardi & Kipp 2008).

Due to the inherent isolation and insularity ofrivers and lakes, freshwater ecosystems often displayhigh rates of endemism (Dudgeon et al. 2006). Eco-systems comprised of highly endemic faunas, such asstreams in the Colorado River Basin, are more likelyto be negatively impacted by invading species, as theinvader often represents a novel predator or competi-tor archetype (Ruesink 2003; Cox & Lima 2006). Inisolation, prey species may not evolve defences (e.g.,behavioural or morphological) against predators notnative to their range (Ruesink 2003), and often expe-rience strong negative effects of introducednon-native predators (Kitchell et al. 1997). Threenon-native fishes in the lower San Rafael River (i.e.,channel catfish, black bullhead and green sunfish)have trophic positions indicative of at least partialpiscivory. While previous studies suggest the large-bodied piscivore trophic niche was occupied season-ally by endangered Colorado pikeminnow in Colo-rado River tributaries (Tyus & Saunders 2000;Bottcher et al. in press), the established populationsof channel catfish, black bullhead and green sunfishpresent a novel, year-round source of predation pres-sure for juveniles of the ‘three species’. Our limitedcatches of juveniles in the river suggest that theyoccur at low abundances, are highly patchy indistribution, or are less susceptible to our samplingmethods. Additional samples with a stick seine inbackwater and slow, marginal habitats also had verylimited catches of juveniles, suggesting that gear biaswas not the cause of the limited catch numbers. Redshiners have been shown to be significant predatorson larval and juvenile stages of Colorado River fishes(Tyus & Nikirk 1990; Tyus & Saunders 2000). Whilewe argue that novel predators likely have substantialnegative effects on juveniles of the ‘three species’,our data show that more juveniles of the ‘three spe-cies’ were captured in the lower San Rafael Riverthan were captured in the upper San Rafael River.This occurrence is likely due to the habitat character-istics of the sampling sites in the upper San Rafael,which, while having extensive complex habitat, didnot have backwater habitats, the preferred habitats ofjuvenile native fishes, available. The contemporarysuite of novel predator archetypes in the riverpresents the potential for substantial impacts on therecruitment success and viability of the ‘threespecies’ (Meffe 1985).In addition to the negative effects of predation

from invasive species, these invaders may also com-pete with native species for resources. Competitionfor resources can be demonstrated by a numericalresponse (Pell & Tidemann 1997), a shift in resourceusage (Werner & Hall 1979; Davey et al. 2006),change in morphology (Crowder 1984), and/or achange in vital rates in the presence of potential com-

9


petitors (Davey et al. 2006). Competition cannot beinferred from stable isotope signatures alone (New-some et al. 2007), as multiple combinations of basalresources and prey items can result in similar isotopicsignatures. Nonetheless, our trophic niche space anal-ysis suggests that the presence of numerous non-native species in the lower San Rafael River has thepotential to increase competition for food resourcesfor the ‘three species’ (DeNiro & Epstein 1978; Zam-brano et al. 2010). The vast majority of the trophicniche space of each of the ‘three species’ is over-lapped by multiple non-native fishes. Furthermore,the trophic niches of the bluehead sucker andflannelmouth sucker are altered in the presence ofnon-native fishes. In the lower San Rafael River, bothspecies demonstrated a more enriched d13C signaturethan in the upper San Rafael River, potentially indic-ative of a shift from riffle to more pool-derivedresources (Finlay et al. 2002). It is important to note,however, that this shift in carbon signature couldresult from competitive exclusion by the non-nativefishes, the scarcity of riffle habitats in the degradedlower San Rafael River, or some combination ofboth. The reduction in the range of nitrogen signa-tures demonstrated by both bluehead sucker and flan-nelmouth sucker may be indicative of nichedisplacement whereby non-native species are limitingthe native fishes’ food resource use to lower trophiclevel items, potentially through exploitative competi-tion over the resource or interference competition rel-egating the native species to less preferred habitats(Douglas et al. 1994).As resources become concentrated into smaller

areas, competition can intensify over an increasinglylimited resource pool (Mills et al. 2004), reducing theenergy available to native species. Competition-basedchanges to a fish’s energy intake (i.e., quantity orquality) can result in less favourable growth patterns,as the fish may not be able to obtain necessaryresources, or may have to increase energy expendi-ture to feed (Mills et al. 2004; Davey et al. 2006).Decreased growth rates can negatively affect fishpopulations, as fecundity (Bagenal 1978), age-at-maturity (Alm 1959), and, often, survival (Quinn &Peterson 1996) are strongly correlated with body size.For example, faster growth rates can enable fish toescape gape-limited predators and avoid starvationduring extreme environmental conditions (Post &Evans 1989; Quinn & Peterson 1996). However, the‘three species’ do not demonstrate lower growth ratesin the lower San Rafael River relative to the upperSan Rafael River, even in the face of lesser resourceabundance (Walsworth 2011) and greater potentialfor competition. Previous studies have suggested thatsource-sink dynamics control the populations of the‘three species’ in the San Rafael River (Bottcher

2009), with the upper San Rafael River and themain-stem Green River providing colonists tothe lower San Rafael sink habitat. The ‘three species’are each highly mobile species that make long dis-tance movements in both downstream and upstreamdirections, either through larval drift or directed adultmigration (e.g., Chart & Bergersen 1992; Robinsonet al. 1998; Compton et al. 2008). Movement of indi-viduals from the Green River into the San RafaelRiver could mask any differences in growth resultingfrom competition. Flannelmouth sucker and blueheadsucker have been shown to grow faster and to largersizes in larger rivers (Sweet et al. 2009). In addition,unidirectional movement downstream over the HattRanch diversion dam, a barrier to upstream move-ment, could mask any fitness-related effects of com-petition with non-native species in the lower river.In addition to disrupting connectivity, dams and

diversions often result in more stable flow regimes.In a static environment, Menge & Sutherland (1987)predicted that the effect of predation outweighs theeffects of competition or physical factors on interme-diate level consumers, such as the ‘three species’,especially in systems with high levels of omnivory.The highly altered contemporary flow regime of theSan Rafael River has led not only to habitat homoge-nisation, but also to frequent low flow and dewater-ing events (Bottcher 2009). Predation should have astronger impact on populations in homogeneous habi-tat (Caroffino et al. 2010) and at times of low flow(Pilger et al. 2010), as the fish are concentrated intosmaller habitat areas with less refuge. The lack ofsignificant growth differences in the San RafaelRiver, coupled with the greater abundances of the‘three species’ in the upper San Rafael River, regard-less of habitat quality, suggest that predation by thenon-native fishes in the lower river has a strongerimpact on the ‘three species’ than does greater com-petition for resources, similar to the findings of others(Pilger et al. 2010).Many of the fish species in our stable isotope anal-

ysis demonstrated elevated carbon signatures relativeto the collector-filterer invertebrates used as a base-line, suggesting the contribution of an alternativecarbon source to the food web. As consumer carbonstable isotope signatures are elevated relative to thecollector-filterers, it is likely that the additionalresources contributing to the food web are detritusand invertebrates from slow-water habitats (Finlayet al. 2002). An alternative basal resource could haveimplications for our interpretation of trophic positionif the nitrogen signature of the primary consumersdemonstrating elevated d13C is different than thesignature demonstrated by the collector-filterers usedin our analysis. However, the results of a previousdesert stream study (Gido et al. 2006) suggest that

10

Walsworth et al.

algae and detritus have similar d15N signatures.Previous studies have demonstrated that each of the‘three species’ is omnivorous, consuming benthicinvertebrates, algae, detritus and in the case of round-tail chub, fish and terrestrial invertebrates (Childset al. 1998; Bezzerides & Bestgen 2002; Quist et al.2006). Thus as collector-filterer insects dominatedour benthic invertebrate samples, we believe that theuse of collector-filterer insects as a baseline is sound.We recognise there are potential limitations to our

inference of both the trophic structure and relativegrowth rates between the upper and lower San RafaelRiver. Previous research suggests using caution whenestimating annual growth from more than the singlemost recent annuli, as this method inherits uncertain-ties due to size-selective mortality (i.e., Lee’sphenomenon; Gutreuter 1987). However, by includ-ing all annuli in the analysis, we were able toestimate growth in years when the populations werenot sampled, or when fin ray sections were notcollected. In addition, an analysis of body conditionrevealed similar results to our size-at-age analysis(Walsworth 2011). Isotopic signatures often varywith fish length and growth rate, as well as over time(Harvey et al. 2002; Vinson & Budy 2011), anotherpotential limitation of our data. However, ontogeneticdiet shifts, if present, would result in more conserva-tive estimates of non-native piscivore trophic posi-tions, as smaller individuals would not yet bepiscivorous. While the data collected for this studycannot explicitly demonstrate predation, greatercompetition or a change in fitness for the ‘three spe-cies’, the weight of evidence from our stable isotopeanalysis, abundance comparisons (T.E. Walsworth &P. Budy in review) and the results of previous studiesthroughout the Colorado River Basin suggest that thenon-native fishes are having substantial impacts onpopulations of the ‘three species’. These negativeimpacts appear to result from the establishment ofnovel predators and competitors in the ecosystem andextensive restructuring of the aquatic community.The reduction or loss of native species in an eco-

system can lead to overwhelming ecosystem leveleffects, including, but not limited to, further speciesloss or reduced ecosystem resilience (Wootton &Downing 2003; Lockwood et al. 2007). Given thehigh densities of non-native fishes in the lower river,the lack of significant increase in native fish densityin complex habitat in the presence of non-nativefishes, and the associated impacts suggested herein,restoration of physical habitat in the lower river mayhave little effect on ‘three species’ populations in theabsence of efforts to minimise the effects of non-native fish species (Walsworth 2011). Improving thephysical habitat template could actually enhance non-native performance and abundance (Bond & Lake

2003). As such, the eradication of non-native fishesmay be a necessary step to ensure the persistence ofthe ‘three species’ throughout their native range,although the effectiveness of such a strategy may bemost effective in combination with restoration of amore natural flow regime (Gido & Propst 2012). Ourresults demonstrate that non-native fish species sig-nificantly alter the food web structure of the SanRafael River, presenting novel predators and compet-itors that threaten the persistence of these endemicand imperilled species. Maintaining connectivity with‘three species’ populations in the upper San RafaelRiver (while avoiding upstream expansion by non-native species) and the main-stem Green River maybe critical to the persistence of the ‘three species’in the lower river (Pulliam 1988; Bottcher 2009;Walsworth 2011).

Acknowledgements

We would like to thank Charles Hawkins and Michelle Bakerfor their thoughtful comments on previous drafts of this manu-script. B. Simcox, J. Remington, K. Wilson, S. Petre, D. Cole,P. Nicholson, R. Chaston, D. Fowler, P. Tuttle, M. Schifiliti,T. Wright, W. Gordon, J. Baker, D. Collins and D. Olsenhelped in the field and laboratory. This research was fundedby United States Bureau of Reclamation Activities to AvoidJeopardy Program, an S.E. and Jessie E. Quinney Fellowship,and the Ecology Center at Utah State University. Additionalsupport was provided by the U.S. Geological Survey – UtahCooperative Fish and Wildlife Research Unit (in kind), theU. S. Bureau of Land Management, and the Utah Division ofWildlife Resources. Two anonymous reviewers provided con-structive comments on previous drafts of this manuscript. Theuse of trade names or products does not constitute endorse-ment by the United States Government. This study was per-formed under the auspices of Utah State University IACUCprotocol 1310.

References

Alm, G. 1959. Connection between maturity, size, and age infishes. Institute of Freshwater Research Drottningholm 40:5–145.

Bagenal, T.B. 1978. Aspects of fish fecundity. In: Gerking,S.D., ed. Ecology of freshwater fish production. New York:Halsted Press, pp 75–101.

Bestgen, K.R. & Propst, D.L. 1989. Distribution, status, andnotes on the ecology of Gila robusta (Cyprinidae) in theGila River drainage, New Mexico. Southwestern Naturalist34: 402–412.

Bezzerides, N. & Bestgen, K. 2002. Status review of roundtailchub Gila robusta, flannelmouth sucker Catostomus latipin-nis, and bluehead sucker Catostomus dicobolus in the Colo-rado River basin. Larval Fish Laboratory Technical Report118. Colorado State University.

Bond, N.R. & Lake, P.S. 2003. Local habitat restoration instreams:constraints on the effectiveness of restoration

11


for stream biota. Ecological Management & Restoration 4:193–198.

Bottcher, J.L. 2009. Maintaining population persistence in theface of an extremely altered hydrograph: Implications forthree sensitive fishes in a tributary of the Green River, Utah.Master’s thesis. Logan: Utah State University.

Bottcher, J.L., Walsworth, T.E., Thiede, G.P., Budy, P. &Speas, D.W. (in press). Frequent Tributary Usage by theEndangered Fishes of the Upper Colorado River Basin:Observations from the San Rafael River, Utah. North Ameri-can Journal of Fisheries Management.

Britton, J.R., Davies, G.D. & Harrod, C. 2010. Trophicinteraction and consequent impacts of the invasive fishPseudorasbora parva in a native aquatic food web: a fieldinvestigation in the UK. Biological Invasions 12: 1533–1542.

Brook, B.W., Sodhi, N.S. & Bradshaw, C.J.A. 2008. Syner-gies among extinction drivers under global change. Trendsin Ecology and Evolution 23: 453–460.

Brouder, M.J. 2001. Effects of flooding on recruitment ofroundtail chub, Gila robusta, in a southwestern river. South-western Naturalist 46: 302–310.

Cabana, G. & Rasmussen, J.B. 1996. Comparison of aquaticfood chains using nitrogen isotopes. Proceedings of theNational Academy of Sciences 93: 10844–10847.

Carey, M.P. & Wahl, D.H. 2010. Native fish diversity altersthe effects of an invasive species on food webs. Ecology 91:2965–2974.

Carey, M.P. & Wahl, D.H. 2011. Determining the mechanismby which fish diversity influences production. Oecologia167: 189–198.

Caroffino, D.C., Sutton, T.M., Elliott, R.F. & Donofrio, M.C.2010. Predation on early life history stages of lake sturgeonin the Peshtigo River, Wisconsin. Transactions of the Ameri-can Fisheries Society 139: 1846–1856.

Case, T.J. 1990. Invasion resistance arises in strongly interact-ing species-rich model competition communities. Proceed-ings of the National Academy of Sciences 87: 9610–9614.

Chart, T.E. & Bergersen, E.P. 1992. Impact of mainstreamimpoundment on the distribution and movements of the resi-dent flannelmouth sucker (Catostomidae: Catostomuslatipinnis) population in the White River, Colorado. South-western Naturalist 37: 9–15.

Childs, M.R., Clarkson, R.W. & Robinson, A.T. 1998.Resource use by larval and early juvenile native fishes inthe Little Colorado River, Grand Canyon, Arizona. Transac-tions of the American Fisheries Society 127: 620–629.

Compton, R.I., Hubert, W.A., Rahel, F.J., Quist, M.C. &Bower, M.R. 2008. Influences of fragmentation on threespecies of native warmwater fishes in a Colorado RiverBasin headwater stream system, Wyoming. North AmericanJournal of Fisheries Management 28: 1733–1743.

Connell, J.H. 1983. On the prevalence and relative importanceof interspecific competition: evidence from field experi-ments. American Naturalist 122: 661–696.

Cox, J.G. & Lima, S.L. 2006. Naivet�e and an aquatic-terres-trial dichotomy in the effects of introduced predators. Trendsin Ecology and Evolution 21: 674–680.

Crowder, L.B. 1984. Character displacement and habitat shiftin a native cisco in southeastern Lake Michigan: evidencefor competition? Copeia 1984: 878–883.

Davey, A.J.H., Turner, G.F., Hawkins, S.J. & Doncaster, C.P.2006. Mechanisms of density dependence in stream fish:exploitation competition for food reduces growth of adultEuropean bullheads (Cottus gobio). Canadian Journal ofFisheries and Aquatic Sciences 63: 597–606.

DeNiro, M.J. & Epstein, S. 1978. Influence of diet on the dis-tribution of carbon isotopes in animals. Geochimica et Cos-mochimica Acta 42: 495–506.

Douglas, M.E., Marsh, P.C. & Minckley, W.L. 1994. Indige-nous fishes of western North America and the hypothesis ofcompetitive displacement: Meda fulgida (Cyprinidae) as acase study. Copeia 1994: 9–19.

Dudgeon, D., Arthington, A.H., Gessner, M.O., Kawabata,Z.I., Knowler, D.J., Leveque, C., Naiman, R.J., Prieur-Rich-ard, A.H., Soto, D., Stiassny, M.L.J. & Sullivan, C.A. 2006.Freshwater biodiversity: importance, threats, status and con-servation challenges. Biological Reviews 81: 163–182.

Eby, L.A., Roach, W.J., Crowder, L.B. & Stanford, J.A. 2006.Effects of stocking-up freshwater food webs. Trends inEcology and Evolution 21: 576–584.

Finlay, J.C., Khandwala, S. & Power, M.E. 2002. Spa-tial scales of carbon flow in a river food web. Ecology 83:1845–1859.

Francis, R.I.C.C. 1990. Back-calculation of fish length: a criti-cal review. Journal of Fish Biology 36: 883–902.

Fortney, S.T., Schmidt, J.C. & Dean, D.J. 2011. Establishingthe geomorphic context for wetland and riverine restorationof the San Rafael River. Final Report. NRCS CooperativeAgreement #68-3A75-4-155.

Gido, K.B. & Propst, D.L. 2012. Long-term dynamics of nativeand nonnative fishes in the San Juan River, New Mexico andUtah, under a partially managed flow regime. Transactions ofthe American Fisheries Society 141(645): 659.

Gido, K.B., Franssen, N.R. & Propst, D.L. 2006. Spatial vari-ation in d15N and d13C isotopes in the San Juan River, NewMexico and Utah: implications for the conservation of nativefishes. Environmental Biology of Fishes 75: 197–207.

Gutreuter, S. 1987. Considerations for estimation and interpre-tation of annual growth rates. In: Summerfelt, R.C., Hall,G.E., eds. Age and Growth of Fish. Ames: Iowa State Uni-versity Press, pp 115–126.

Harvey, C.J., Hanson, P.C., Essington, T.E., Brown, P.B. &Kitchell, J.F. 2002. Using bioenergetics models to predictstable isotope ratios in fishes. Canadian Journal of Fisheriesand Aquatic Sciences 59: 115–124.

Johnson, B.M., Martinez, P.J., Hawkins, J.A. & Bestgen, K.R.2008. Ranking predatory threats by nonnative fishes in theYampa River, Colorado, via bioenergetics modeling. Trans-actions of the American Fisheries Society 28: 1941–1953.

Jones, C.M. 2000. Fitting growth curves to retrospective size-at-age data. Fisheries Research 46: 123–129.

Kaehler, S. & Pakhomov, E.A. 2001. Effects of storage andpreservation on d13C and d15N signatures of selected marineorganisms. Marine Ecology Progress Series 219: 299–304.

Kelly, B., Dempson, J.B. & Power, M. 2006. The effects ofpreservation on fish tissue stable isotope signatures. Journalof Fish Biology 69: 1595–1611.

Kitchell, J.F., Schindler, D.E., Ogutu-Ohwayo, R. & Reinthal,P.M. 1997. The Nile perch in Lake Victoria: interactionsbetween predation and fisheries. Ecological Applications 7:653–664.

12

Walsworth et al.

Layman, C.A., Arrington, D.A., Montana, C.G. & Post, D.M.2007. Can stable isotope ratios provide for community-widemeasures of trophic structure? Ecology 88: 42–48.

Lockwood, J.L., Hoopes, M.F. & Marchetti, M.P. 2007. Inva-sion Ecology. Oxford: Blackwell.

Marvier, M., Kareiva, P. & Neubert, M.G. 2004. Habitatdestruction, fragmentation, and disturbance promote inva-sion by habitat generalists in a multispecies metapopulation.Risk Analysis 24: 869–878.

Meffe, G.K. 1985. Predation and species replacement inAmerican southwestern fishes: a case study. SouthwesternNaturalist 30: 173–187.

Menge, B.A. & Sutherland, J.P. 1987. Community regulation:variation in disturbance, competition, and predation in rela-tion to environmental stress and recruitment. American Nat-uralist 130: 730–757.

Mills, M.D., Rader, R.B. & Belk, M.C. 2004. Complexinteractions between native and invasive fish: the simulta-neous effects of multiple negative interactions. Oecologia141: 713–721.

Minagawa, M. & Wada, E. 1984. Stepwise enrichment of 15Nalong food chains: further evidence and the relation betweend15N and animal age. Geochimica et Cosmochimica Acta48: 1135–1140.

Minckley, W.L. & Deacon, J.E. 1968. Southwestern fishesand the enigma of “endangered species”. Science 159:1424–1432.

Newsome, S.D., Martinez del Rio, C., Bearhop, S. & Phillips,D.L. 2007. A niche for isotopic ecology. Frontiers in Ecol-ogy and the Environment 5: 429–436.

Olden, J.D., Poff, N.L. & Bestgen, K.R. 2006. Life-historystrategies predict fish invasions and extirpations in the Colo-rado River Basin. Ecological Monographs 76: 25–40.

Olsson, K., Stenroth, P., Nystrom, P. & Graneli, W. 2009.Invasions and niche width: does niche width of an intro-duced crayfish differ from a native crayfish? FreshwaterBiology 54: 1731–1740.

Pell, A.S. & Tidemann, C.R. 1997. The impact of two exotichollow-nesting birds on two native parrots in savannah andwoodland in eastern Australia. Biological Conservation 79:145–153.

Pilger, T.J., Gido, K.B. & Propst, D.L. 2010. Diet and trophicniche overlap of native and nonnative fishes in the GilaRiver, USA: implications for native fish conservation. Ecol-ogy of Freshwater Fish 19: 300–321.

Post, J.R. & Evans, D.O. 1989. Experimental evidence ofsize-dependent predation mortality in juvenile yellow perch.Canadian Journal of Zoology 67: 521–523.

Pulliam, H.R. 1988. Sources, sinks, and population regulation.American Naturalist 132: 652–661.

Quinn, T.P. & Peterson, N.P. 1996. The influence of habitatcomplexity and fish size on over-winter survival and growthof individually marked juvenile coho salmon (Oncorhynchuskisutch) in Big Beef Creek, Washington. Canadian Journalof Fisheries and Aquatic Sciences 53: 1555–1564.

Quist, M.C., Bower, M.R. & Hubert, W.A. 2006. Summerfood habits and trophic overlap of roundtail chub and creekchub in Muddy Creek, Wyoming. Southwestern Naturalist51: 22–27.

R Development Core Team. 2011. R: a language and environ-ment for statistical computing. Vienna, Austria: R Founda-

tion for Statistical Computing. ISBN 3-900051-07-0, URLhttp://www.R-project.org.

Reissig, M., Trochine, C., Queimalinos, C., Balseiro, E. &Modenutti, B. 2006. Impact of fish introduction on plank-tonic food webs in lakes of the Patagonian Plateau. Biologi-cal Conservation 132: 437–447.

Ricciardi, A. & Kipp, R. 2008. Predicting the number of eco-logically harmful exotic species in an aquatic system. Diver-sity and Distributions 14: 374–380.

Ricciardi, A. & Rasmussen, J.B. 1999. Extinction rates ofNorth American freshwater fauna. Conservation Biology 13:1220–1222.

Ricker, W. 1975. Computation and interpretation of biolog-ical statistics of fish populations. Technical Report Bulle-tin 191, Bulletin of the Fisheries Research Board ofCanada.

Robinson, A.T., Clarkson, R.W. & Forrest, R.E. 1998. Dis-persal of larval fishes in a regulated river tributary. Transac-tions of the American Fisheries Society 127: 772–786.

Romare, P. & Hansson, L.A. 2003. A behavioral cascade:top-predator induced behavioral shifts in planktivorous fishand zooplankton. Limnology and Oceanography 48: 1956–1964.

Ross, S.T. 1986. Resource partitioning in fish assemblages: areview of field studies. Copeia 1986: 352–388.

Ruesink, J.L. 2003. One fish, two fish, old fish, new fish:which invasions matter? In: Kareiva, P., Levin, S.A., eds.The Importance of Species. Princeton University Press:Princeton, pp 161–178.

Sarakinos, H.C., Johnson, M.L. & Vander Zanden, M.J. 2002.A synthesis of tissue-preservation effects on carbon andnitrogen stable isotope signatures. Canadian Journal of Zool-ogy 80: 381–387.

Schoener, T.W. 1983. Field experiments on interspecific com-petition. American Naturalist 122: 240–285.

Serrano, O., Serrano, L. & Mateo, M.A. 2008. Effects of sam-ple pre-treatment on the d13C and d18O values of livingbenthic foraminifera. Chemical Geology 257: 218–220.

Shea, K. & Chesson, P. 2002. Community ecology theory as aframework for biological invasions. Trends in Ecology andEvolution 17: 170–176.

Simberloff, D. & Von Holle, B. 1999. Positive interactions ofnonindigenous species: invasional meltdown? BiologicalInvasions 1: 21–32.

Sousa, R., Nogueira, A.J.A., Gaspar, M.B., Antunes, C. &Guilhermino, L. 2008. Growth and extremely high produc-tion of non-indigenous invasive species Corbicula fluminea(M€uller, 1774): Possible implications for ecosystem func-tioning. Estuarine, Coastal, and Shelf Science 2: 289–295.

Sweet, D.E., Compton, R.I. & Hubert, W.A. 2009. Age andgrowth of bluehead suckers and flannelmouth suckers inheadwater tributaries, Wyomng. Western North AmericanNaturalist 69: 35–41.

Tyus, H.M. & Karp, C.A. 1990. Spawning and movements ofrazorback sucker, Xyrauchen texanus, in the Green RiverBasin of Colorado and Utah. Southwestern Naturalist 35:427–433.

Tyus, H.M. & Nikirk, N.J. 1990. Abundance, growth, and dietof channel catfish, Ictaluris punctatus, in the Green andYampa Rivers, Colorado and Utah. Southwestern Naturalist35: 188–198.

13


Tyus, H.M. & Saunders, J.F. III 2000. Nonnative fish controland endangered fish recovery: Lessons from the ColoradoRiver. Fisheries 25: 17–24.

UDWR, Utah Division of Wildlife Resources. 2006. Conser-vation and management plan for three fish species in Utah:addressing needs for roundtail chub Gila robusta, blueheadsucker Catostomus discobolus, and flannelmouth suckerCatostomus latipinnis. Salt Lake City: Utah Division ofWildlife Resources.

Unmack, P.J. & Fagan, W.F. 2004. Convergence of differen-tially invaded systems toward invader-dominance: time-lagged invasions as a predictor in desert fish communities.Biological Invasions 6: 233–243.

USDI, Fish and Wildlife Service 1994. Endangered and threa-tened wildlife and plants; determination of critical habitatfor the Colorado River endangered fishes; razorback sucker,Colorado squawfish, humpback chub, and bonytail chub(Final Rule). Federal Register 59: 13374.

Vander Zanden, M.J. & Rasmussen, J.B. 1999. Primary con-sumer d13C and d15N and the trophic position of aquaticconsumers. Ecology 80: 1395–1404.

Vinson, M.R. & Budy, P. 2011. Sources of variability and com-parability between salmonids stomach contents and isotopicanalyses: study design lessons and recommendations. Cana-dian Journal of Fisheries and Aquatic Sciences 68: 137–151.

Walker, C.A. & Hudson, M. 2004. Surveys to determine thecurrent distribution of the roundtail chub, flannelmouthsucker, and bluehead sucker in the San Rafael drainage, dur-ing 2003. Salt Lake City: Utah Division of WildlifeResources. Unpublished report.

Walsworth, T.E. 2011. Analysis of food web effects of non-native fishes and evaluation of restoration potential for theSan Rafael River Utah. Master’s thesis. Logan: Utah StateUniversity.

Werner, E.E. & Hall, D.J. 1979. Foraging efficiency and hab-itat switching in competing sunfishes. Ecology 60: 256–264.

Werner, E.E., Gilliam, J.F., Hall, D.J. & Mittelbach, G.G.1983. An experimental test of the effects of predation riskon habitat use in fish. Ecology 64: 1540–1548.

Woodford, D.J. 2009a. Evidence of source-sink metapopula-tions in a vulnerable native galaxiid fish driven by intro-duced trout. In: Riverscape-mediated effects of introducedtrout on native galaxiid fishes in New Zealand. Ph.D. disser-tation. Christchurch: University of Canterbury, pp. 14–35.

Woodford, D.J. 2009b. Biotic and abiotic drivers of co-occur-rence between introduced trout and native alpine galaxias(Galaxias paucispondylus) across New Zealand riverscapes.In: Riverscape-mediated effects of introduced trout on nativegalaxiid fishes in New Zealand. Ph.D. dissertation. Christch-urch: University of Canterbury, pp. 38–59.

Wootton, J.T. & Downing, A.L. 2003. Understanding theeffects of reduced biodiversity: A comparison of approaches.In: Kareiva, P., Levin, S.A., eds. The Importance of Species.Princeton: University Press Princeton, pp 161–178.

Zambrano, L., Valiente, E. & Vander Zanden, M.J. 2010.Food web overlap among native axolotl (Ambystoma mex-icanum) and two exotic fishes: carp (Cyprinus carpio) andtilapia (Oreochromis niloticus) in Xochimilco, Mexico City.Biological Invasions 12: 3061–3069.

14

Walsworth et al.

longer food chains and crowded niche space: effects of

Documents